This is a group for academics, students, industry practitioners and all interested persons to discuss new topics related to music information retrieval, music technology and machine learning applied to music.
We're excited to have Aleksey Tikhonov discuss his work on using Variational Auto Encoders (VAEs) and Adversarially Constrained Autoencoder Interpolations (ACAIs) to generate novel drum patterns. He will be speaking at the offices of IDAGIO on Monday, June 24.
This talk is about my recent project on a drum patterns latent space exploration. I've collected a large dataset of drum patterns, then used a neural network approaches like VAE and ACAI to map them into a latent explorable space with some recognizable genre areas, so one can explore this space and use it to generate some new patterns.
Aleksey Tikhonov, a lead analyst in Yandex, co-founder of creAIted.com (http://creaited.com).
Has 17+ years experience in IT and ML. Works for Yandex since 2010 as a data analyst, a researcher and team lead.
Last years participated in several projects on neural network music and poetry generation.